https://www.sensetime.com
https://en.wikipedia.org/wiki/SenseTime
SenseTime and IMPALA are reality based data recoding/analysis systems that parse what is recorded in real time or stored real time recording to parse and reveal information and knowledge derived from what is seen and heard.
https://www.quora.com/Where-did-the-phrase-who-are-you-going-to-trust-me-or-your-lying-eyes-originate
Ironically originated in a movie!
https://www.reuters.com/article/us-china-monitoring-insight/from-laboratory-in-far-west-chinas-surveillance-state-spreads-quietly-idUSKBN1KZ0R3
The size of SenseTime: August 2018
https://techcrunch.com/2018/08/08/ai-giant-sensetime-leads-199m-investment-in-chinese-video-tech-startup/
My blog entry yesterday preceding this one focused on IMPALA. A new AI learning algorithm.
In the course of this blog I have from time to time focused on Natural Language Processing and Hardware/Software system recognition of spoken human language as well as vocal and transcribed to text expression of Natural Language. That technology can be applied to analysis of any and all historically recorded spoken language to parse the aggregated spoken record of an individual person it for meaning and derive information and knowledge at the intelligence gathering level. More socially important people probably tend to have had more of what they say recorded than normal. Historically true too a proved by the best text record example. What might that example reveal if parsed by AI?
SenseTime acquired Moviebook. An entertainment sector business.
Entertainment Sector: It produces sells and retains in a data base recorded entertainment that has a variety of genres from science fiction to documentary.
That would be an interesting data base to which AI could be applied to reveal information and knowledge. Applied not only to what is spoken (and probably transcribed to subtitle text more recently) but to what is graphically presented.
This is where IMPALA comes into the relationship.
IMPALA can learn from looking at the real world and learn behaviors from what it sees. Interesting use of the term actors and agents in the IMPALA algorithm! It can look at entertainment monies and learn from them probably just as well as looking at the visual and spoken real world.
What might IMPALA learn to recognize and the learn, parse and present as information to us for our knowledge about ourselves and what entertains us. What entertains us tells us something about collective selves as well as our personal selves if what we view (Netflix for example)......very interesting way of looking in a mirror at our society as well as the person in the mirror.
In the 60's a college course involving watching movies was popular. Actually just another variation of reading a book to earn but a more popular mode of learning for examining what is presented and its meaning...for whatever worthy purpose.
How Fast could IMPALA "view" a set of movies or all movies ever made to categorize them and then intelligently present to us what it sees?
IMPALA...Tell me about this movie. IMPALA tell me about al movies ever made.
IMPALA tell me about guns and violence in movies. About love. Now there is something that challenges the intelligence of a thinking machine!
IMPALA could answer those questions? Maybe the new test of machine intelligence is if it can tell us why we love movie so much!
Movies are not the real time real world but shadows on the wall. A script that was written about long ago. The person that walked out of that cave came back to those still viewing shadows on the wall and told them about the outside that projected only shadows on the wall.
Those imprisoned did not get it! Stayed in their seats to watch the shadows on the wall that was their Bohemian Rhapsody reality.
Truth is undergoing a stress test.
Will AI set the standard of Truth that takes us beyond the standard we individually and collectively set for ourselves?
No comments:
Post a Comment